Recent advances in the knowledge of the complexity and specificity of neural circuits suggest that understanding how neural circuits generate perception and behavior will be nearly impossible with presently available techniques. Because different neuron types involved in distinct subcircuits are intermingled, and even neighboring neurons of the same type differ in their connectivity and function (DeAngelis et al., J. Neurosci., 19:4046-64, 1999; Song et al., PLoS Biol., 3:e68, 2005; Ohki et al., Nature, 433:597-603, 2005; Yoshimura et al., Nature, 433:868-73, 2005), methods are required which can reveal the connections both of specific cell types and especially of single neurons.
Historically, the identification of neural networks has involved labor-intensive and technologically challenging methods, such as examining sections of neural tissue with light or electron microscopy (Gilbert, Ann. Rev. Neurosci., 6:217-47, 1983; Douglas and Martin, Ann. Rev. Neurosci., 27:419-51, 2004; Gray, Nature, 183(4675):1592-3, 1959; Timofeeva et al., J. Neurosci., 25(40):9135-43, 2005), simultaneous patch recording from pairs of neural cells (Mercer, et al., Cereb. Cortex, 15(10):1485-96, 2005), and photostimulation-based mapping of connections in brain slices (Callaway and Katz, Proc. Natl. Acad. Sci. USA, 90(16):7661-5, 1993; Shepherd and Svoboda, J. Neurosci., 25(24):5670-9, 2005; Zarrinpar & Callaway, J. Neurophysiol., 95(3):1751-61, 2006). None of these methods permit a wholesale way of identifying neurons that are connected either to some other cell group or, especially, to a single cell.
Transsynaptic tracers (DeFalco et al., Science, 291:2608-13, 2001; Zou et al., Nature, 414:173-9, 2001; Braz et al., Proc. Natl. Acad. Sci. USA, 99:15148-53, 2002; Maskos et al., Proc. Natl. Acad. Sci. USA, 99:10120-5, 2002), out of all the available techniques, might appear to offer a solution to this problem: By introducing a tracer into a particular cell or cell type, synaptically connected cells should be labeled by the tracer and therefore be identifiable as those in synaptic contact with the starting cells in question. Transsynaptic tracers can be introduced into particular neurons or populations of neurons using a variety of methods (Maskos et al., Proc. Natl. Acad. Sci. USA, 99(15):10120-5, 2002; DeFalco et al., Science, 291(5513):2608-13, 2001; Zou et al., Nature, 414(6860):173-9, 2001; Braz et al., Proc. Natl. Acad. Sci. USA, 99(23):15148-53, 2002; Ruda & Coulter, Brain Res., 249(2):237-46, 1982; Evinger & Erichsen, Brain Res., 380(2):383-8, 1986; Ugolini et al., Brain Res., 422(2):242-56, 1987; Ugolini et al., Science, 243(4887):89-91, 1989; Kuypers & Ugolini, Trends Neurosci., 13(2):71-5, 1990; Ugolini, J. Comp. Neurol., 356(3):457-80, 1995; Kelly & Strick, J. Neurosci. Methods, 103(1):63-71, 2000; Aston-Jones & Card, J. Neurosci. Methods, 103(1):51-61, 2000). However, no such method has been sensitive enough to label cells synaptically connected to a single cell of origin.
Moreover, due to their dependence on cellular machinery for transport to and across synapses (Vercelli et al., Brain Res. Bull., 51:11-28, 2000), traditional transsynaptic tracers cross different synapses at different rates: The more hardware servicing in a given connection, the more efficiently it will be traversed by a traditional tracer. As shown schematically in FIG. 1, tracer that accrues in transsynaptically labeled cells will begin spreading in turn to the cells that are connected to them and in fact can label the most strongly connected of these even before weakly connected synaptic partners of the starting population (Ugolini et al., Brain Res., 422:242-56, 1987; Ugolini et al., J. Comp. Neurol., 356:457-80, 1995). The result of asynchronous transsynaptic transfer is an'inescapable ambiguity in the number of synapses crossed by traditional transsynaptic tracers.
No technique to date has been capable of identifying en masse neurons that are connected directly to a primary neuron (or population of primary neurons) of interest. The best available tools, transsynaptic tracers, cross multiple synapses and are unable to distinguish weak direct connections from strong indirect ones (Ugolini et al., Brain Res., 422:242-56, 1987; Ugolini et al., J. Comp. Neurol., 356:457-80, 1995). Furthermore, no tracer has proven potent enough to label any connected neurons whatsoever when starting from a single cell. Thus, there is a need for a tracer that crosses one synaptic step, to cells directly connected to the starting cell or cell population, and then stops, unable to spread beyond them to indirectly connected cells.